If an abend situation occurs in TMSXTEND, it will probably recover by itself. Look for messages TMSXT41I and TMXST49I. If the machine that TMSXTEND is running on goes down, there will be some manual recovery necessary.
The first scenario is when the TMC being extended is shared by more than 1 CPU. This is the easiest situation to recover from. Simply run TMSXTEND with PARM=RECOVER on any other CPU that shares the same TMC and AUDIT.
The second scenario is when the TMC being extended is only accessed from 1 (one) CPU, and that CPU goes down. After it uses IPL and CAS9 has run, but BEFORE the TMSINIT job has been run, you may be required to perform some manual intervention. If the TMSXT30I message had been issued, but NOT the TMSXT31I message OR TMSXT21I messages, then you must manually rename TMC.O to TMC before running TMSINIT. In this situation, after TMSINIT has been run you should also run TMSXTEND with a PARM=RECOVER to finish the cleanup of CTL#4.
In the same scenario, if the TMSXT30I message had not yet been issued by TMSXTEND, then there is no recovery necessary and you may simply run TMSINIT.
Also in this same scenario, if the TMSXT31I AND TMSXT21I messages had both already been issued, there is also no recovery necessary and you may simply run TMSINIT.
|
Copyright © 2013 CA Technologies.
All rights reserved.
|
|